Software Patents: A Time for Change?

Over the years, nearly all limits to patentable subject matter in the U.S. have been removed by the specialized patent court without input from Congress, empirical evidence to justify its decisions, or consideration of values and principles outside of the patent system. Problems related to software patents have sparked calls for patent reform here in the U.S. and have driven efforts to reject legislation authorizing US-style software patents in Europe.

The persistent, and now resurgent, controversy over software patents calls for a reexamination of the issues in light of experience and new research. What has been learned about the impact on software development, innovation, and competition?  How has this changed as software patents have become increasingly common and used in unanticipated ways?  And what should now be done to change software patent policy here in the U.S.?

***

The controversy over software patents in the U.S. has continued for over 40 years. The 1966 report of the President’s Commission on the Patent System recommended against patents for computer programs on pragmatic grounds – that they could not readily be examined for adherence to the basic criteria of novelty, utility, and nonobviousness. Today, software patents are considered a done deal in the U.S., thanks to the decisions of the specialized patent court of appeals. Indeed, this court has abolished all limits on patentable subject matter, including methods of doing business, teaching and learning, and communication of information and ideas. Vast numbers of patents on software and other abstract subject matter have issued, and the companies holding these patents are often fiercely protective, and assertive, of what they see as their property. Yet the problems and the controversy persist.

Today, patent reform has become a major issue, but it is caught up in inter-industry divisions. The greatest demand for reform comes from software and related sectors, where the problems are especially acute. In examining how different sectors experience the patent system, the 2003 FTC report, To Promote Innovation, found that it was viewed most negatively in the areas of software and electronic commerce. This led to a recommendation to “consider possible harm to competition – along with other possible benefits and costs – before extending the scope of patentable subject matter.” In the course of the FTC’s joint hearings with the DOJ, it became apparent that computer technology and semiconductors experienced many of the problems experienced by software developers, although perhaps not to the same degree.

Internationally, the situation is very different. Many countries retain provisions excluding patents on computer programs, although the U.S. has been demanding acceptance of software patents in negotiations for bilateral trade agreements. The European Patent Office has been issuing tens of thousands of software patents despite the European Patent Convention’s exclusion of computer programs. Pressures to validate EPO practice lead to proposed “Directive on Computer-Implemented Inventions” that provoked intense public debate and culminated in a historically unprecedented rejection of the directive on its second reading in the European Parliament.

On the other hand, the U.S. and European experiences can be viewed as two sides of the same problem because of the one-size-fits-all nature of the patent system. If software is patentable, as in the U.S., it may suffer disproportionately under a regime optimized for high-cost pharmaceuticals. Europe confronts the question of whether software (or certain kinds of software) should be locked into an ill-fitting set of laws from which there may be no escape.

The U.S. Patent & Trademark Office held hearings on the policy aspects of software patents in 1994. At the time, the views of patent lawyers, computer companies, and pureplay software companies were nearly uniform within categories. But software patents were a recent phenomenon 1994. Today we have had another 12 years of practical experience with software patents, as well as the emergence of a growing body of empirical research in both the U.S. and Europe – something the Commissioner noted was missing in his introduction to the San Jose hearing in 1994. In the interim, virtually all subject matter barriers have been eliminated by decisions of the specialized Court of Appeals for the Federal Circuit. Patents are routinely granted for financial services, business models and practices, teaching and learning methods, communication and presentation of information, production and use of diagnostic information, and algorithms and other tools for producing scientific knowledge. Today, most of these abstract processes depend on software to varying degrees.

Because it is so easy to create functionality in software, it appears especially vulnerable to the overpatenting that leads to the phenomena of patent flooding, thickets, and minefields. Thus patents on software enable companies to develop portfolios that protect their market positions, but this may make it difficult for newcomers to break into product markets. The high information costs and uncertainty surrounding software patents also creates a fertile breeding ground for “trolls.” In this opaque environment, there are many opportunities to take advantage of surprise and the vulnerability of producers and users.

Software is exceptionally vulnerable to patent ambush because of the extraordinary importance of standards. The networked nature of software, especially in global Internet and World Wide Web makes collaborative development of standards essential to the introduction of new technology and building confidence in new applications and services. Yet success in standards take-up combined with the economies of scale in software distribution become sources of extreme vulnerability at the hands of trolls who can now assert inadvertently patents against whole industry segments and users as well as producers.

Because software has become so pervasive and deeply embedded in social and economic activity, software patents also raise policy issues that are not presented by patents on more traditional subject matter. In particular, software patents increasingly impinge on processes of communication, teaching, learning, publishing, information access, and scientific research. All of which are closely tied to freedom of expression and inquiry and other values that have no role in the administration, adjudication, and enforcement of patents.

Software patents also affect competition and innovation in fields that have traditionally not been subject to patents and where patent protection has never been sought. This includes not only financial services, advertising, marketing, fund-raising, sales, insurance, accounting, and other services. Clearly, patents create winners and losers within these areas of activity, but what is the net benefit or cost?

This conference will examine the changing debate over software patents from a variety of perspectives. What has been learned about software patents over the past 20 years? How have they shaped the behavior of developers, vendors, and users? How has the debate been reformulated by the growth of the Internet and World Wide Web? The trend toward open standards? The growth of open source software? How are patents affecting education and human interaction?

Finally, how can the problems be addressed? How should they be addressed in light of changing technology and evolving business practice? What can be done about the growing conflicts between patents and standards? Is patent reform a possible answer? Or are there structural and institutional obstacles that cannot be overcome? Where do we go from here in addressing the problems – not just in the U.S. but internationally as well?